Stability of Rosmarinic Acid in Aqueous Extracts from Different Lamiaceae Species after in vitro Digestion with Human Gastrointestinal Enzymes

Research on biologically active compounds, such as plant phenolics, represents a big scientifi c challenge due to a great number of identifi ed substances, their chemical diversity as well as their huge biological potential, important for the maintenance of healthy balance in our body. Undoubtedly, rosmarinic acid is one of the most studied phenolic compounds, exhibiting diff erent biological activities: antioxidant, anti-infl ammatory, antimutagenic, antibacterial, antiviral, cytotoxic on human breast cancer cells, neuroprotective in human neurons, and nephroprotective (1–8). In plants, rosmarinic acid is formed from amino acids phenylalanine (caff eic acid part of rosmarinic acid) and tyrosine (3,4-dihydroxyphenyllactic acid part) (9). It was isolated for the fi rst time from Rosmarinus offi cinalis and its structure was characterized as an ester of caff eic acid and 3,4-dihydroxyphenyllactic acid (10,11). According to some authors (12,13), rosmarinic acid cannot be used as chemotaxonomic marker for diff erentiation of plant families due to its occurrence in various plant families such as Lamiaceae, Boraginaceae, Blechnaceae and Asteraceae. However, Lamiaceae species such as Satureja montana L., Thymus vulgaris L. and Origanum majorana L. are known as rich sources of rosmarinic acid (14–16). ISSN 1330-9862 scientifi c note


Introduction
Research on biologically active compounds, such as plant phenolics, represents a big scientifi c challenge due to a great number of identifi ed substances, their chemical diversity as well as their huge biological potential, important for the maintenance of healthy balance in our body.Undoubtedly, rosmarinic acid is one of the most studied phenolic compounds, exhibiting diff erent biological activities: antioxidant, anti-infl ammatory, antimutagenic, antibacterial, antiviral, cytotoxic on human breast cancer cells, neuroprotective in human neurons, and nephroprotective (1)(2)(3)(4)(5)(6)(7)(8).
In plants, rosmarinic acid is formed from amino acids phenylalanine (caff eic acid part of rosmarinic acid) and tyrosine (3,4-dihydroxyphenyllactic acid part) (9).It was isolated for the fi rst time from Rosmarinus offi cinalis and its structure was characterized as an ester of caff eic acid and 3,4-dihydroxyphenyllactic acid (10,11).According to some authors (12,13), rosmarinic acid cannot be used as chemotaxonomic marker for diff erentiation of plant families due to its occurrence in various plant families such as Lamiaceae, Boraginaceae, Blechnaceae and Asteraceae.However, Lamiaceae species such as Satureja montana L., Thymus vulgaris L. and Origanum majorana L. are known as rich sources of rosmarinic acid (14)(15)(16).
The majority of published studies on rosmarinic acid are focused on methods of its detection in various plant extracts (17,18), its seasonal variations in selected plants (19) as well as its specifi c biological activity in diff erent models (cell cultures, rats, in vitro methods) (5,8,20,21).Therefore, there is a lack of studies on the stability of rosmarinic acid under gastrointestinal conditions using human gastrointestinal enzymes and on the infl uence of plant matrix on its stability rate.In that sense, in vitro models that mimic human physiology are recognized as simple, inexpensive and reproducible tools to study digestive stability of compounds (22).In addition, in vitro digestion methods are generally based on the use of commercial digestive enzymes, while studies with human gastrointestinal enzymes are still very rare.
Thus, the aim of this study is to determine the infl uence of digestion phase (gastric and duodenal), temperature and pH on the stability of rosmarinic acid.In this study human gastrointestinal juices (gastric and duodenal) collected from healthy donors are used instead of commercial gastrointestinal enzymes.The stability rate of rosmarinic acid is measured by HPLC-DAD technique.

Plant material
Plant materials used in this study were dry leaves of Thymus vulgaris L. (thyme), Satureja montana L. (winter savory), and Origanum majorana L. (lemon balm).The above--mentioned samples are commercially available, in form of tea-like preparations, and were purchased from a local herbal pharmacy (Suban d.o.o., Strmec Samoborski, Croatia).Pretreatment of the plant material included its homogenization using high speed grinder for 1 min.

Preparation of aqueous extracts
Pulverized plant materials (2.5 g) were extracted with distilled water (25 mL).To obtain bett er extraction yield, instead of conventional procedure, ultrasonic bath-assisted extraction was used.Flasks with suspensions were immersed in ultrasonic water bath heated to 50 °C.The extraction time was 1 h.Aft er cooling, the samples were fi ltered and centrifuged for 5 min at 1800×g.The extraction of each sample was performed in triplicate.The obtained extracts of the same plant species were combined into the fi nal extract that was used in further experiments.

Collection of human digestive juices
Human gastric and duodenal juices were collected from four donors (two male and two female) without known gastrointestinal pathology, and who were not taking acid secretion inhibitors or antibiotics.Gastric and duodenal juices were aspirated through the endoscope.Eight hours before the procedure, all liquid or food intake was ceased.For each patient, 3 mL of initially aspired juice were discarded and the remaining amount was collected in a sterile tube, which was centrifuged to remove mucus and cell debris.In order to reduce interindividual variations, batches of pooled gastric and intestinal juices were prepared and then stored at -20 °C until use.The approval for the collection of digestive juices was obtained from the Ethics Committ ee of the University Hospital Centre Split, Croatia.

Determination of enzymatic activity of collected juices
The procedure described by Almaas et al. (23) was used to determine enzymatic activity of the prepared pooled human gastric juice samples.Pepsin activity was measured using 2.5 % solution of bovine haemoglobin.The solution was prepared in 0.2 M phosphate buff er (pH=7.6)and then acidifi ed (to pH=3) using H 2 SO 4 .In order to determine the human duodenal juice activity, casein solution (1 %) dissolved in 0.2 M phosphate buff er (pH=7.6)was used.A volume of 500 μL of prepared protein solutions was incubated with 5, 20 or 50 μL of gastrointestinal juice.The digestion reactions were stopped with the addition of 1 mL of 10 % trichloroacetic acid (TCA).Samples were measured spectrophotometrically at 280 nm.One unit (U) of enzyme activity is defi ned as the amount of enzyme that causes the absorbance change of 1 between the blank and the sample, aft er 20 min at 37 °C.

In vitro digestion process
A two-phase digestion procedure was performed according to the method described by Furlund et al. (24).Gastric and intestinal digestion phases were performed at 37 °C, in shaking bath (180 rpm).The volume of digestive juice corresponding to 1 U of enzymatic activity was 20 μL of human gastric juice and 25 μL of human duodenal juice.Before in vitro digestion procedure, the concentration of rosmarinic acid in the prepared aqueous extracts from selected plants was normalized (471.48 mg/L).For digestion, 4 mL of aqueous extracts were used.The pH of the samples was adjusted to pH=2.5 using 1 M HCl for gastric phase, and to pH=7.5 using 2 M NaOH for intestinal phase.The concentration of human juices used for this assay was 20 U per g of plant material for gastric and 62.4 U per g of plant material for intestinal phase.In order to determine the infl uence of diff erent concentrations of juices on the stability of rosmarinic acid, following concentrations were used: 5, 10, and 20 U per g of plant material for gastric phase, and 15.6, 20, and 62.4 U per g of plant material for intestinal phase.The incubation period of gastric phase was 30 min, while aliquots of intestinal samples were collected aft er 60 and 120 min of intestinal phase.Enzymatic reactions were stopped on ice and the samples were stored at -20 °C until analyses.All digestion processes were run in duplicate.Stability rate of rosmarinic acid represents the ratio of its concentration before in vitro digestion and aft er gastric or intestinal digestion phases.

HPLC-DAD detection of rosmarinic acid
The rosmarinic acid was analyzed by a direct injection of the extracts, previously fi ltered through a 0.45-μm pore size membrane fi lter (Macherey-Nagel GmbH & Co. KG, Düren, Germany).Chromatographic separation was performed using HPLC instrument with Agilent 1260 quaternary LC Infi nity system (Agilent Technologies, Santa Clara, CA, USA) equipped with diode array detector (DAD), an automatic injector and ChemStation soft ware.The separation of compounds was performed on a Nucleosil 100-5C18 (250 mm×4.6 mm, i.d. 5 μm) column (Macherey-Nagel GmbH & Co. KG).The solvent composition and the used gradient conditions were described previously by Generalić Mekinić et al. (13).For gradient elution, mobile phase A contained 2 % of acetic acid (T.T.T., Sveta Nedelja, Croatia) in water, while solution B contained 2 % of acetic acid in acetonitrile (BDH Prolabo, VWR International, Lutt erworth, UK).The used elution program was as follows: from 0 to 3 min 2 % B, from 3 to 25 min 30 % B, from 25 to 35 min 80 % B, and fi nally for the last 5 min again 0 % B. The fl ow rate was 1.0 mL/min and the injection volume was 24 μL.Detection was performed with UV/Vis-photo diode array detector (Agilent Technologies) by scanning from 250 to 300 nm.Identifi cation of rosmarinic was carried out by comparing retention times and spectral data with those of the authentic standards at 280 nm.The quantifi cations of rosmarinic acid were made by the external standard method.Working rosmarinic acid standard solutions were prepared by diluting the stock solution to yield five concentrations in a range from 175 to 700 mg/L.Quantitative determination was carried out using the calibration curves of the standard (y=31.834x,R 2 =0.99).Quantitative determination was based on peak area from HPLC analyses and from the mass concentration of the compound.The results were expressed in mg per mL of extract, as mean value±standard deviation (N=2 replicates).

Statistical analysis
Statistical analysis was performed using GraphPad InStat3 soft ware (GraphPad Soft ware Inc., San Diego, CA, USA).The relationship between the obtained parameters was described using Pearson's correlation coeffi cient r.Diff erences at p<0.05 were considered to be statistically signifi cant.

Results and Discussion
Lamiaceae species are known as a rich source of rosmarinic acid (14)(15)(16).In this study, three diff erent plants from the Lamiaceae family were used: Thymus vulgaris L., Origanum majorana L. and Satureja montana L. Prepared aqueous extracts contained diff erent concentrations of rosmarinic acid, measured by HPLC-DAD (Fig. 1).Therefore, before the two-step in vitro digestion, the concentration of rosmarinic acid was normalized in all prepared aqueous extracts so that the concentration measured in thyme (471.48 mg/L) was taken as primary concentration before in vitro digestion.Table 1 shows the concentration and the stability rate of rosmarinic acid in plant extracts in comparison with pure rosmarinic acid aft er the gastric digestion phase (pH=2.5).The duration of gastric phase was 30 min because liquids with a low protein concentration are considered to have very short transit time in the stomach (25).Rosmarinic acid was the most stable in Origanum majorana L. extract (14.10 %) in comparison with Satureja montana L. extract (5.8 %) and Thymus vulgaris L. extract (0.8 %) aft er digestion phase.However, the stability rate of pure rosmarinic acid was signifi cantly higher (30.77%) than in rosmarinic acid from plant mate-rial.Incubation temperature (37 °C) did not aff ect the stability of rosmarinic acid, which is in accordance with the results of other authors (26,27).On the other hand, the incubation period of 30 min in the acidic medium (pH=2.5)decreased its concentration by more than 50 % (Table 1).The obtained results are not in accordance with those presented by Dinis et al. (28), who concluded that rosmarinic acid in Mentha species was stable at low pH under simulated gastrointestinal conditions using commercial enzymes (pepsin and pancreatin).Contrary to commercial enzymes, human digestive juices consist of a variety of enzymes, inhibitors and bile salts that collectively contribute to the digestion of food sample (29).In human gastric juices the following components have been detected: pepsin, trypsin, gastricsin, bile, small peptides and protein fragments (30).Duodenal juice contains pancreatic and intestinal enzymes such as proteolytic enzymes, intestinal lipases, enterokinase, trypsinogen, chymotrypsinogen, and amylase (31).
Results of the stability determination of rosmarinic acid aft er intestinal digestion phase are shown in Table 1.Aliquots of digested samples were collected aft er 60 and 120 min of intestinal phase.Duration time of 120 min for intestinal phase was chosen according to in vivo results obtained by Troost et al. (32).The stability aft er intestinal digestion phase of pure rosmarinic acid diff ers completely from the stability of rosmarinic acid in plant extracts (Table 1).In comparison with its primary concentration in plant extracts, rosmarinic acid was almost completely degraded aft er intestinal digestion phase.Aft er 120 min of intestinal digestion, the highest concentration was detected in Origanum majorana L. extract (30.8 mg/L).The obtained results are not in line with the results by Dinis et al. (28), who reported that pancreatin (commercial mixture of amylase, lipase and protease) did not provoke the degradation of rosmarinic acid in Mentha species.Also, Bel--Rhild et al. (33) did not observe any hydrolysis of rosmarinic acid aft er the passage of rosemary extract through the gastrointestinal tract model (using commercial digestive enzymes).Putative reasons for such discrepancies could be explained by two factors: the diff erence in in vitro digestion procedure between commercial digestive enzymes and human gastrointestinal enzymes, and the matrix eff ect -the rosmarinic acid can behave diff erently in Mentha species and/or Rosmarinus offi cinalis extract in comparison with other plant species containing it.In contrast to rosmarinic acid in aqueous plant extracts, pure rosmarinic acid showed very high resistance to digestion by intestinal human juices (its stability rate was approx.78 % aft er 120 min of intestinal phase).In addition, according to the results presented in Table 1, the degradation of rosmarinic acid aft er intestinal phase is not infl uenced by the pH.At pH=7.5 the rosmarinic acid is almost completely stable.Table 2 shows the infl uence of diff erent concentrations of human gastric and duodenal juices on the stability of rosmarinic acid in the prepared aqueous extract from Satureja montana L. In order to investigate the infl uence of the concentration of human gastrointestinal juices, three diff erent concentrations were tested: 5, 10 and 20 U per g of plant material of gastric juices, and 15.6, 20 and 62.4 U per g of plant material of duodenal juices.The obtained results showed that rosmarinic acid stability depends on the concentration of human gastrointestinal juices (the lowest the concentration of digestive juice, the highest the stability of rosmarinic acid; Table 2).
Higher stability rate of rosmarinic acid aft er in vitro intestinal digestive phase may be also related to its ability to react with pancreatic enzymes that inhibit their activity   (34).It was reported that rosmarinic acid-derived quinones react with amino acid side chains and free thiol groups of the enzyme (35).In addition, the interaction between phenolic acids and fl avonoids has been reported and the presence of some fl avonoids such as luteolin and apigenin enhances the stability rate of rosmarinic acid (36,37).According to the available data, the highest content of luteolin and apigenin among the three selected Lamiaceae plants was detected in Origanum majorana aqueous extracts, in which the rosmarinic acid stability was the highest aft er both, gastric and intestinal phases (38).

Conclusions
There is a lack of information about the gastrointestinal stability of rosmarinic acid.Results of this study showed several interesting observations on digestive stability of rosmarinic acid using in vitro digestion with human gastrointestinal enzymes.The huge diff erence was found in the stability of pure rosmarinic acid and rosmarinic acid in aqueous extracts from diff erent plant material belonging to Lamiaceae family.Pure rosmarinic acid was more stable aft er duodenal phase (78.64 %) than during gastric phase (30.77 %).Acid medium greatly reduced the stability of rosmarinic acid (>50 %), while the eff ect of the incubation temperature and slightly alkaline medium were not signifi cant.There were diff erences in the stability of rosmarinic acid among diff erent plant extracts.In our study, rosmarinic acid showed the highest stability after both digestive phases in Origanum majorana aqueous extract.In addition, the concentration of human gastrointestinal juices aff ected the stability of rosmarinic acid in both digestive phases.Finally, the obtained results showed lower gastrointestinal stability of rosmarinic acid than in other studies using commercial digestive enzymes.

Fig. 1 .
Fig. 1.The investigated Lamiaceae species and their HPLC-DAD chromatograms with the signed peak of the dominant compound (rosmarinic acid) and its structure.The rosmarinic acid concentration in Satureja montana L., Origanum majorana L. and Thymus vulgaris L. extracts was 1579.56,821.06 and 471.48 mg/L, respectively.Photos taken from Wikimedia Commons database

Table 1 .
Concentrations and stability rates of rosmarinic acid in aqueous extracts of Thymus vulgaris L., Satureja montana L. and Origanum majorana L. aft er in vitro gastric and intestinal digestion

Table 2 .
Concentrations and stability rates of rosmarinic acid in aqueous extract of Satureja montana L. aft er in vitro gastric and intestinal digestion in relation to diff erent concentrations (U per g of plant material) of human gastric juices (HGJ) and human intestinal duodenal juices (HDJ)The concentration of rosmarinic acid in undigested aqueous extract was 471.48 mg/L.Results are expressed as mean value±standard deviation